The overall purpose of this dissertation research is to provide detailed documentation of the meanings of long-term care as perceived by Korean American elderly and examine their preferred long-term care arrangement in the case of physical and functional limitation and cognitive impairment. The specific aim of this research is to conduct a cross-sectional study of the expected process and outcome of long-term care decision-making among Korean American elderly by examining the role of culture, values, acculturation, physical and functional health, family relationship, and the knowledge of community and institutional long-term care decision-making. A cross-sectional survey involving face-to-face interviews with a probability sample of 180 community-dwelling Korean American elderly in Los Angeles County is proposed. A systematic two- stage area/cluster sampling procedure will be used to draw a sample for this study. The interview will be conducted by trained bilingual/bicultural interviewers. Using survey questionnaire with open-ended and closed-ended questions, this research will collect information in the following areas: demographic characteristics (age, gender, education, marital status, living arrangement, and immigration experience), antecedent characteristics (physical health history, the level of functional limitation, and socioeconomic status), mediating factors (family relationship, adherence to traditional values and cultural norms, acculturation, locus of control, and the knowledge of community or institutional long-term care services), and the expected process and preferred outcome of long-term care decision-making. Both qualitative and quantitative method of inquiry will be directed to data analysis. Qualitative analysis method such as content analysis is proposed to analyze data obtained from open- ended questions by identifying, coding, and categorizing the primary patterns in open-ended responses to four different hypothetical illness scenarios. For quantitative data, descriptive and explanatory analyses will be conducted to describe univariate, bivariate, and multivariate relationships among variables under study. Specifically, explanatory analyses such as binary and multinomial logistic regressions and latent class analysis will be directed to examine overall relationships between dependent variables with demographic, antecedent, and moderating variables.